When I wrote about Matt Forte’s top-five upside, I created projections for the rest of the Jets skill players. I thought I’d share those today. Along the way, I’ll show you how we create projections at RotoViz.
To generate projections, we have a proprietary Projection Machine that forces us to account for team-level factors – pass-to-run ratio, average scoring margin, pace – as well as coordinator and player-level factors. That makes it hard to project just one player, but that’s a good thing. What I mean is that each player’s performance is dependent on the players around him, so you really have to look at the whole team in order to come up with a number for one guy.
We also have a subscriber-level Projection Machine that will be up and running soon, and that you can use to make your own projections. With that in mind, think of this as a bit of a tutorial as well as an actual (albeit early) team projection.
Team Level Settings
Right off the bat you’ll notice that my settings are slightly different than I used in my Matt Forte article. In that article, I was trying to be very conservative, so I wouldn’t overstate the case for Forte. But this time around, I’m going with what I think is actually most likely to happen. These settings affect every other projection, so in that sense they’re the most important settings to consider.
The average score differential, per play, that’s expected for 2016. Once we have good Vegas lines, we can use those to impute the average margin. For now I’ve just set it to league average.
The Jets were essentially neutral last year, and I’ve left them that way in my projection.
How many plays, based on game situation, do the Jets run vs. the average team? I’ve set them well above average at 1.75 plays per game above. But as you can see from the underlying graph, that’s actually quite a bit lower than their 2015 pace. So I’m projecting some reversion to the mean while still being optimistic about their overall pace. Last year they ran 1,074 plays; I’m projecting 1,053 in 2016.
I’m assuming Ryan Fitzpatrick re-signs with the Jets. Note that I’ve set his sack rate very low – in the bottom quartile of all QBs. But that’s well above his rate for last season. Fitzpatrick’s sack rate was very low – for him – last year, so I expect it to increase closer to his career rate of 5.6 percent.
Since the Jets did sign Forte – and Bilal Powell, and Khiry Robinson – I’ve marked down Fitzpatrick’s rushing attempts, but still left them in the top quartile. In his past four years, Fitzpatrick has never averaged less than 11 percent of team attempts.
I project Fitzpatrick’s interception rate to be basically the same as last year, which was basically league average, which is basically what he’s been for his career.
Ryan Fitzpatrick Projection
|QB||Fitzpatrick, Ryan NYJ QB||578.70||362.24||0.63||3964.86||24.57||15.91||48.83||163.09||1.90||268.63||268.63|
Nothing outrageous. Attempts and yards are in line with last year’s performance. I did cut his TDs back a bit, in part because I think Brandon Marshall will be less productive and in part because I think his 31 TDs a year ago were a bit of a high water mark for him. The QB Sim App suggests he’ll throw fewer TDs/game, as did most of his comparable QBs in their N+1 season.
Wide Receiver Settings
Maybe the most interesting thing is that I’m projecting Quincy Enunwa as the WR3, not Devin Smith. The WR Sim App likes him more, and Smith is still rehabbing from a torn ACL. Plenty of time to change this, but that’s where I’m at right now.
I cut Marshall’s target percentage from 28 percent last year down to 25. I also lowered his yards/target by 0.68 and his TD rate by 1.5 percent. I think he’ll have another excellent year, but regression is very likely; 20 of 25 comps declined in their N+1 season. I’ve still got him pegged as league average or better compared to all WR1s.
I left everything the same as 2015 except TD rate. Last year’s nine percent rate seems really tough to duplicate. On the other hand, Decker has almost always had a very high TD rate.
I’d feel comfortable bumping his rate up a bit, but I’ll stick with seven percent for now.
Wide Receiver Projections
|WR1||Marshall, Brandon NYJ WR||143.91||87.79||1151.31||13.11||9.35||0.00||0.00||5.32||0.00||259.04||215.15|
|WR2||Decker, Eric NYJ WR||126.64||76.75||985.29||12.84||8.87||0.00||0.00||4.85||0.00||228.47||190.09|
|WR3||Enunwa, Quincy NYJ WR||34.54||18.24||233.14||12.78||1.04||0.00||0.00||4.80||0.00||47.77||38.65|
That’s 30 fewer targets for Marshall, which doesn’t seem like much, but it works out to 84 fewer points (5.3/game) than he scored last year. Still enough to be a WR1 (last year’s PPR WR12, Calvin Johnson, had 263 points), but a sizable step down in production.
Decker’s numbers also take a mild step back. The targets, receptions, and yards are very similar to last year (132/80/1027) but I project many fewer TDs. I want to go back in and bump that number up, but there’s a couple reasons I won’t right now. First is the fact that, last year, only 10 WRs in the entire league posted double digit receiving TDs. It seems unlikely it will happen again. Second is the fact that the Jets signed Forte and Khiry Robinson. It seems likely that some TDs will find their way to those guys.
Tight End Settings
Last year Jets tight ends earned 25 targets. Combined. For the whole season. And they posted a league-worst -0.74 per-target efficiency rating. Brutal. So I think my eight percent allocation of targets to whomever is the Jets TE1 is generous. It’s just below the 25th percentile, but as you can see in the graph, it’s way more than the position earned last year. Nothing I’ve seen so far suggests Jace Amaro is on track to be a major part of the Jets plans offensively. Again, this could change as OTAs, training camp, etc. unfold. But for now I see nothing to like here.
Tight End Projections
|TE1||Amaro, Jace NYJ TE||46.05||27.63||276.31||10.00||0.00||55.26||41.45|
|TE2||Body, Warm NYJ TE||19.57||12.43||120.21||9.67||1.14||31.26||25.05|
Running Back Settings
Why am I doing this when I just wrote an article about Forte’s outlook? Because in that article, I configured the settings very conservatively, to see how much potential downside there was to Forte’s 2016 outlook. This time I’m using numbers that are still reasonable, but more accurately reflect what I think will really happen.
Throughout this exercise, I’ve been drawing on the NFL Career Graphs App to inform my estimates. In this case, I’ve projected Forte’s yards per carry, TD rate, catch rate, and yards per target in line with his past couple seasons. I’ve given him a slightly below league average rushing workload for an RB1, and a bullish, top-quartile share of pass targets.
I bumped up his market share of rush attempts a bit, and really boosted his market share of targets. Everything else is in line with his recent performance.
I really don’t see much pass game work for Robinson, but I did give him a very high rushing TD Rate. I do think it makes sense that he gets more short yardage/goal line work than the other backs, so that’s how I tried to reflect that.
Running Back Projections
|RB1||Forte, Matt NYJ RB||221.96||910.03||4.44||74.83||56.13||449.01||8.00||1.48||227.52||199.46|
|RB2||Powell, Bilal NYJ RB||79.90||351.58||1.12||86.35||63.03||526.72||8.36||1.61||167.21||135.70|
|RB3||Robinson, Khiry NYJ RB||66.59||266.35||4.00||13.82||9.60||74.01||7.71||0.29||69.35||64.55|
I’ve got Forte down for just nine more carries and 12 more rushing yards than he saw last season, so it’s not like I’m projecting massive rejuvenation or anything. I also have him down for just one more pass target per game than last season. As for Powell, I still see him as a borderline top-20 RB. It’ll be interesting to follow the ADP of these two players. While I’m confident in Forte and think there are situations where he’s a good fantasy target, Powell may turn out to be a better value, based on his cost. Finally, barring injury to the other two backs – and the ability to finally stay healthy himself – I don’t see significant value for Robinson.
We’ll be doing team projections all spring and summer, so look forward to more posts. We won’t always go into this level of detail, but I wanted to give you a sense of the thought processes that go into making projections. We’ll also use our projections to power our recommendations in the Best Ball and Dynasty ADP apps. Finally, you’ll be able to apply your own thought process and craft your own projections via the Projection Machine.